Movement Detection Using Image Processing · Detection of moving object in an image sequences is crucial issue of moving video. A very common difficulty in the detection of moving

International Journal of Science and Research (IJSR) ISSN (Online): 2319-7064

Index Copernicus Value (2013): 6.14 | Impact Factor (2013): 4.438

Volume 4 Issue 2, February 2015 www.ijsr.net

Licensed Under Creative Commons Attribution CC BY

Movement Detection Using Image Processing

Shweta S. Kshirsagar1, P. A. Ghonge2

1, 2 Mumbai University, Yadavrao Tasgaonkar Institute of Engineering & Technology, Bhivpuri Road, Karjat, India

Abstract: In this paper, we present an algorithm for moving object detection from video sequences. Here we propose new method for

the movement detection i.e. Background subtraction method .We establishes reliable background updating model by using threshold

method to detect complete moving object. By using filtering background disturbance and noise will be eliminated. The used method will

quickly and accurately runs and fit for real-time detection.

Keywords: background subtraction; background model; moving object detection.

1. Introduction

Detection of moving object in an image sequences is crucial

issue of moving video. A very common difficulty in the

detection of moving object is the presence of ego motion.

We solve it by computing the dominant motion . Visual

surveillance is a very active research area in computer vision.

The use of this concept in surveillance for security, fight

against terrorism and crime. The main task includes motion

detection , object classification ,tracking activity. The

detection of moving objects in video streams is the first

relevant step of information extraction in many computer

vision applications. Many organizations and institutions

needs to secure their facilities thus need to use security and

surveillance systems that are equipped with latest

technology. Intelligent video sensors were developed to

support security systems to detect unexpected movement

without human intervention. The important information of

move , location , speed and any desired information of target

from the captured frames can be taken from the camera and

can be transferred to the analysis part of the system.

Movement detection is one of these intelligent systems to

which detect and tracks moving targets. There are different

methods to detect moving objects but these methods having

some limitations for real time application. Due to this reason,

in this paper we use background subtraction method, which

is more suitable for real time application which gives

accurate result.

In this paper we use method in which single static camera to

capture video for that captured video we convert in to frame

and apply background subtraction method with updating

background.

2. Method Used For Detection of Moving Object

The inter-frame difference method modeled by mixture of

Laplacian distribution and Gibbs random field is used for

describing the label set [1].Representing moving images as

set of overlapping layers. An image coding involves three

parts encoder, representation and decoder . The most

common image coding system used today based on Low

level image processing concept like DCT‟s [2]. Tracking

moving object in cluttered scene is also done by background

modeling. Two processes involved in this method are

technique to extract an initial background model when

background situation changes at later time [3]. The

knowledge segmentation and edge detection is used in this

modeling. Different types of moving objects share

similarities but also express differences in terms of their

dynamic behavior and the nature of movement [4,5].

Different method used for movement detection .frame

subtraction method is suitable for variety of dynamic

environment but generally it is difficult to get complete

outline of moving object, as result the detection of moving

object is not accurate. Large quantity of calculation,

sensitivity to noise, poor anti-noise performance makes

Optical Flow method not suitable for real-time demanding

occasions. The background subtraction method is to use the

difference method of the current image and background

image to detect moving objects, with simple algorithm, but

very sensitive to the changes in the external environment and

has poor anti-interference ability. The Background method is

very sensitive to the changes in the external environment and

has poor anti-interference ability.

3. Advanced Method Used For Movement Detection

This paper describes a real-time system for human detection,

tracking and motion Analysis. The system is an automated

video surveillance system for detecting and monitoring

people in both indoor and outdoor environments. Detection

and tracking are achieved through several steps: First, we

design a robust, adaptive background model that can deal

with lightning changes, long term changes in the scene and

objects occlusions. This model is used to get foreground

pixels using the background subtraction method. Afterwards,

noise cleaning and object detection are applied, followed by

human modeling to recognize and monitor human activity in

the scene such as human walking or running. Techniques

Used:

Motion Detection

Tracking

Human Model

Surveillance

Motion Analysis

Image Processing

3.1 Application of Background Subtraction Method

The background subtraction method is the common method

of motion detection. It is a technology that uses the

Paper ID: 24021504 2329





difference of the current image and the background image to

detect the motion region, and it is generally able to provide

data included object information. The key of this method lies

in the initialization and update of the background image. The

effectiveness of both will affect the accuracy of test results.

Therefore, we use an effective method to initialize the

background, and update the background in real time.

Input video is given as input which converted in to frame .

From these frames images are separated. From these images

initial background image is constructed. This image divided

in to two images current frame image and background frame

image. After separation background subtraction method

applied to detect moving object for next frame background

updated.

A. Background image initialization There are many methods to get initial background image

.Time average method can not deal with image shadow

problems. While the method of taking the median from

continuous multiframe can resolve this problem simply and

effectively. So the median method is selected for background

initialization.

B. Background Update For the background model can better adapt to light changes,

the background needs to be updated in real time, so as to

accurately extract the moving object. The camera is fixed;

the background model can remain relatively stable in the

long period of time. At this stage we can effectively avoid

the unexpected phenomenon of the background, such as the

sudden appearance of something in the background which is

not included in the original background. Moreover by the

update of pixel gray value of the background, the impact

brought by light, weather and other changes in the external

environment can be effectively adapted.

C. Moving Object Extraction To extract moving object dynamic threshold method is

suitable by using dynamically changed threshold values

according to lighting changes of the two images .This

method effectively suppressed the effect of light changes.

Figure 3.1: Background subtraction

D. Reprocessing As the complexity of the background, the difference image

obtained contains the motion region, additionally it also

contains large number of noise. Therefore, noise needs to be

removed. To remove this noise we use median filter with the

3 X 3 window and filters out some noise. After the median

filter, in addition the motion region, includes not only body

parts, but also may include moving cars, flying birds,

flowing clouds and swaying trees and other nobody parts.

Morphological methods are used for further processing.

Firstly, corrosion operation is taken too effectively to filter

out non-human activity areas. Secondly, using the expansion

operation to filter out most of the non-body motion regions

while preserving the shape of human motion without injury.

After expansion and corrosion operations , some isolated

spots of the image and some interference of small pieces are

eliminated ,and we get more accurate human motion region.

E. Extraction of Moving Human Body After median filtering and morphological operations, some

accurate edge regions will be got, but the region belongs to

the moving human body could not be determined. Also it is

seen that while moving object appears shadow will also

appear in some region of scene. It will affect to accuracy of

moving object detection. By using vertical and horizontal

projection to detect the height of motion region. This will

help to eliminate the impact of the shadow up to certain

degree. Human body detection is to identify the

corresponding part of human from the moving region. But

the extracted moving region may correspond to different

moving objects, such as pedestrians, vehicles and other such

birds, floating clouds, the Swaying tree and other moving

objects. Hence we use the shape features of motion regions

to further determine whether the moving object is a human

being. Judging criteria are as follows: (1) The object area is

larger than the set threshold (2) The aspect ratio of the object

region should conform to the set ratio. If these two

conditions are met, the moving object is the moving human

body, or is not a human body.

4. Methodology

This system follows following stages to track moving object

Input video Actually input video is taken from Real time camera is also

possible and also we can able to apply pre defined video

(MATLAB support only .avi format video)

Frame Separation In this project we need to convert input video to frame‟s by

using MATLAB. Combination of frame‟s is called as video.

Each and every video have no of frames, by using that value

we can take the no of frame‟s.

Current Image & Background Image After converting video to frames first frame (image) is called

background image, and other then first image are current

image and these images are not similar at the time of moving

object detection video. And also these current images are

similar at the time of non moving object detection[20].

∆𝑡=λ.1

(𝑀×𝑁) 𝐹 𝑖, 𝑗 − 𝐵(𝑖, 𝑗) 𝑀−1

𝑗=0 𝑁−1𝑖=0 ----(4.1)

Then,

Dif_frm (x,y)= 1 𝐹𝑘 𝑥, 𝑦 − 𝐵𝑘 − 1(𝑥, 𝑦) > t+∆𝑡------(4.2)

Background Subtraction Background subtraction means we just subtract the current

image and background image, and the current image is

updated in each and every time, background image is

constant. by using these technique we can easily find the

Paper ID: 24021504 2330





moving object. For ex. background image pixel – current

image pixel

If the output is 0 means no moving object detected.

If the output is 1 means moving object detected.

Preprocessing & shape analysis By using this technique we can identify the shape of the

moving object ( if there is any variation in background image

and current image ).otherwise no moving object is detected (

if there is no variation in background image and current

image ).

Working flow of project 1) First we are taking one video file as input then we are

going to convert video to “n number of frames” by using

MATLAB. and save the frames in any format with proper

name. Here we do 30 frames of taken video.

2) In frame separation we use image resizing .

3) In that video first frame is consider as Background image

(say as previous image), other than first image we

consider as current image. (and background image is

constant but current image is different based on iteration

and video). By subtracting background image from

current image we get difference image. Before

background subtraction we have to convert every current

image from RGB to Grey scale.

4) For this difference image we apply thresholding as

follows Difference image >35=1 otherwise 0

For dynamic thresholding,

Difference image >100=1.Otherwise 0

5) After background subtraction we are going to update the

current image by using loop Fn. 6) Then we are applying the median filter for noise

suppression.

7) Using Morphological functions we track object. If the

background image and current image is similar means no

moving object detected or moving object detected. For

more than one object we use different outbox of different

colors to track them.

8) In validation by comparing Ground Truth image with our

output image we estimate the parameters like MSE ,

Entropy, Precision, PSNR etc.

Implementation Technical Requirement

Software Requirements Platform: Windows 7

Programming Language: MATLAB Version 7.9.0.529

(R2009b)

Hardware Requirements: Main processor: Intel Core i3

processor 2.30GHz.

RAM: 4 GB Hard Disk: 160 GB

5. Result and Conclusion

The problems associated with the background-based moving

object detection techniques are mainly due to the variations

of ambient lighting. In this paper we introduced an effective

moving object detection algorithm based on method

background updating technique under conditions where

illumination cannot be controlled.

We have also introduced the process of background updating

in this technique, at every frame. That is why we introduced

an improvement of this method that makes it function well

even when there is a moving object detected in the scene

(when background updating is locked out which makes the

algorithm susceptible to illumination changes in that period).

The new described algorithm was applied on three more

video sequences and it also showed very good and promising

results. One of the colored sequences had two moving

objects. The proposed method succeeded to completely

successfully detect the both moving objects in the scene

independently of the luminance conditions. We showed in

this paper the results obtained from the one of the used

sequences that was the most convenient for the

representation. The proposed algorithm, invariant to external

luminance changes, has been tested under various lighting

conditions, artificially simulated on the computer and with

the moving object brighter and darker than the background,

and satisfactory and promising results have been achieved.

The key of this method lies in the initialization and update of

the background image.

The effectiveness of both will affect the accuracy of test

results.

Therefore, this paper uses an effective method to initialize

the background, and update the background in real time.

6. Conclusion A real computer vision system able to model a stationary

object background or a movement object background in

cluttered environment has been presented. The proposed

system is based on the modeling of the structure of the scene.

The quality of the detection is improved when the

background is highly texture .Therefore in our future works

we will use this modeling method in our object tracking

system for tracking rigid and non –rigid movement object . It

will be used for to tracking multiple non-rigid movement

objects in a cluttered scene.

Paper ID: 24021504 2331





7. Future Scope

Future work will be directed towards achieving the following

issues:

Object classifications, Better understanding of human motion

not only vehicle, including segmentation and tracking of

articulated body parts .Improved datalogging and retrieval

mechanisms to support 24/7 system operations .Better

camera control to enable smooth object tracking at high

zoom, in case, video is vibrating Video stabilization

algorithm is required. Acquisition and selection of “best

views” with the eventual goal of recognizing individuals in

the scene.

Figure 4.1: Input Data

Figure 4.2: Background subtracted image

Figure 4.3: Tracking object

Figure 4.4: Output result of video

Figure 4.5: Graphical representation of movement of object

as displacement Vs time

References [1] Adaptive detection of moving objects using multiscale

technique by N. Paragios

[2] Representing moving images with layer by John

Y.A.Wang

[3] Background modeling for tracking object movement by

YueFeng

[4] Exploring movement similarity, analysis of moving

object by Somayesh Dodge

[5] Self organizing approach to background subtraction for

visual surveillance application byLucia

Maddalena&AlfradoPetrisino.

[6] A.R.J.Francois,G.G.Medioni, Adaptive colour

background modeling colour backgroundsegmentation

of video stream. IEEE 2012 Transaction.

[7] Prati, I. Mikic, M. Trivedi, R. Cucchiara,Detecting

Moving Shadow : formulationAlgorithm and evaluation.

[8] T. Horprasert, D. Harwood, L.S. Davis, A statistical

approach for real-time Robustbackground subtraction

and shadow detection.

[9] Du-Ming Tsai and Shia-Chih Lai, "Independent

Component Analysis Based BackgroundSubtraction for

Indoor Surveillance," Image processing IEEE

Transaction a volume 18issue, 1Jan 2009

[10] NiuLianqiang and Nan Jiang, "A moving objects

detection algorithm based on improvedbackground

subtraction,"Intelligent Systems Design and

Paper ID: 24021504 2332





Applications, 2008.ISDA'08.Eighth International

conference on volume3, 26-28 Nov 2008.

[11] T. Aach, A. Kaup, and R. Mester.Statistical model-based

change detection in moving video. Signal Processing,

31:165{180, 1993.

[12] M. Abdel-Mottaleb and A. Elgammal. Face detection in

complex environments from color images. In IEEE

International Conference on Image Processing, Kobe,

Japan, 1999.

[13] D. Comaniciu, V. Ramesh, P. Meer, “Kernel-Based

Object Tracking”, IEEE Trans. on Pattern Analysis and

Machine Intelligence, vol.25, No. 5, 2003.

[14] D. Comaniciu, V. Ramesh, P. Meer, “Real-Time

Tracking of Non-Rigid Objects using Mean Shift”, IEEE

Conf. on Computer Vision and Pattern Recognition

(CVPR „00), vol.2, 2000.

[15] Th. Gevers, “Color in Image Search Engines”, Survey

on color for image retrieval from Multimedia Search. M.

Lew, Springer Verlag, 2001.

[16] D. Comaniciu and P. Meer, “Mean Shift: A robust

approach toward feature space analysis”, IEEE Trans. on

Pattern Analysis and Machine Intelligence, vol.24, No.

5, 2002.

[17] A. Jepson, D. Fleet, and T. El-Maraghi, “Robust online

appearance models for visual tracking”, in Proc. IEEE

Conf. on Computer Vision and Pattern Recognition,

Hawaii, vol. I, 2001.

[18] V. Ferrari, T. Tuytelaars, and L. V. Gool, “Real-time

affine region tracking and coplanar grouping”, in Proc.

IEEE Conf. on Computer Vision and Pattern

Recognition, Hawaii, vol. II, 2001.

[19] C. Sminchisescu, and B. Triggs, “Covariance scaled

sampling for monocular 3D body tracking”, in Proc.

IEEE Conf. on Computer Vision and Pattern

Recognition, Hawaii, vol. I, 2001.

[20] C. Wren, A. Azarbayejani, T. Darrell, and A. Pentland,

“Pfinder: Real-time tracking of the human body”, IEEE.

Author Profile Shweta S. Kshirsagar is pursuing M.E (Electronics &

Telecommunication), B.E (Eletronics& Telecommunication)

Lecturer inYadavraoTasgaonkar Institute Of Engineering &

Technology, India .

Prof. P. A Ghonge, M.E (Electronics & Telecommunication).

HOD Electronics Department,YadavraoTasgaonkar Institute Of

Engineering & Technology,India. Prof. P.A Ghonge is having 15

years teaching experience.

Paper ID: 24021504 2333

Documents

Movement Detection Using Image Processing · Detection of moving object in an image sequences is crucial issue of moving video. A very common difficulty in the detection of moving